Preamble
Our air conditioner broke this summer. It was installed the spring after we moved into the house, which made it 9 years old. It wasn’t that the breakage was expensive (it was just the starting capacitor), but rather that it was the second breakage in as many years (previous year was a compressor replacement, and although the part was still under warranty, the labour cost, refrigerant, and diagnostic cost all come out of my pocket, to the tune of nearly $600). So we decided to replace it, and started to wonder about using a heat pump instead. Our decision to replace was justified – we continued to use the A/C even with the broken starting capacitor (we just started it manually) and by the end of the summer, “something else” had broken in the A/C, since it made an awfully loud racket when operating and was no longer able to cool the house fully. We were happy to say good riddance to it.
The furnace was a consideration too. Our furnace was in good condition, but we think no newer than circa 2007 and quite possibly older. There was no imminent need to replace it, but it was probably some 50-75% of the way through its lifespan. Since it was reasonable to replace the full hvac system, this gave us an opportunity to switch to a heat pump system. You may never easily pry my ICE-based car from my hands, but a heat pump would at least leave us feeling some moral license for the car.
Available Heat Pump Options
A major first consideration for us was the amount of space available for the exterior unit (heat pumps are composed of an exterior unit and one or more interior units). Based on the space available in the location where we wanted to put the exterior unit, there were really just two choices that made sense — a large Mitsubishi or a small Mitsubishi. The large unit is advertised as a 3 ton unit (it is actually a bit more) and is eligible for what was, at the time, a $5000 subsidy. It is suitable only for homes with fully ducted heating/cooling. The smaller unit is advertised as a 2.5 ton unit (it is actually a bit less) and is not eligible for any subsidy. The advertised tonnage is misleading and the smaller unit is actually almost 30% less capacity than the larger unit. However, the smaller unit supports both ducted and ductless heating/cooling, which gave us the option to resolve an air conditioning issue in the master bedroom — older homes such as ours do not have an upstairs air return, and so air conditioning often struggles to cool the top half of the house. A window unit in the master bedroom would be cheap enough, but wasn’t really what we wanted.
We ultimately selected the smaller 2.5 ton Mitsubishi, with a 2.5 ton ducted air handler and a 0.5 ton ductless head to provide supplemental cooling for the top half of the house. This exceeds the rated capacity of the outdoor unit but our hvac contractor confirmed with directly with the Mitsubishi representative that this was okay (the ductless head would only be used for cooling in summer, and 2 tons of cooling is sufficient for the ducted portion of our hvac). The smaller unit was a bit cheaper even accounting for the $5000 subsidy (be careful to compare after-tax prices, since the $5000 subsidy doesn’t get grossed-up to cover tax). As well, if we bought the large 3 ton Mitsubishi, we would no longer be eligible for a subsidy to replace our windows. Since my wife is eager to replace the windows for aesthetic reasons (and I admit they are a bit ugly), the loss of the subsidy on the windows increases the effective cost of the 3 ton Mitsubishi. Subsidies are complex.
For full disclosure, Mitsubishi also had another option for ducted-only configurations that was nearly the same as the high capacity unit and would have made sense if we were, for some reason, ineligible for the $5000 subsidy. We eliminated this from consideration early on, however, since it only made sense in the non-subsidy scenario. (Again, subsidies are complex….)
So what I’m wondering is how the two choices compare, now that we have some experience with the unit.
An important caveat here — I really have no expertise in hvac systems, so all my calculations are really just estimates. I’ve taken the published numbers and derated them a bit based on the assumption that real-life doesn’t work as well as in a laboratory setting. Where necessary, I’ve loosely extrapolated or interpolated.
If you want someone who actually has a bit more clue for what they are talking about, try this blog post here: https://talkintrashwithuhn.com/2022/01/28/mike-at-home-heat-pump-at-20c/
Required Heating
First of all, supposedly a modern-built Energystar-rated three story townhome requires about 4.5 kW heating at -20C. That’s about 1.5 ton, which is to say … not very much. My house is a 70 year-old, under-insulated, 1.5 story detached home. For me, the small Mitsubishi maxes-out at around -10C to -15C, depending on wind and the target interior temperature – so that’s about 2 tons, or 7kW (I’m assuming we don’t achieve its 8kW max capacity since presumably Mitsubishi measured 8kW under ideal conditions, whereas in reality performance varies with exterior humidity levels, and probably a half-dozen other variables). To get to -20C, I figure I need about another 2kW (a very rough guess – take this with a decent size grain of salt), so probably around 9kW. We keep our indoor temp set a bit colder than most people (avg around 18C; cooler at night, warmer in the morning) – not to save energy, but just because we like it that way during winter (less shock when transitioning outside, helps keep the kids in bed in the morning, etc). Someone else might plausibly need an extra 1-2kW, if they’re targeting 20C or 21C indoors.
I’m using kW here, since that’s easy to work with when calculating hydro costs. Roughly $0.10/kWh hydro cost, weighted, and 3.5kW per ton. Data sheets are often in BTU/hr, so helpful to remember 12000 BTU per ton, too.
Heating Available
The heat pump capacity drops-off after -15C. Interestingly, the drop-off is greater for the large Mitsubishi than for the small Mitsubishi. The real difference, however, is actually in the size of the backup coils that they support. The 2.5 ton indoor air handler maxes-out at a 10kW (rated) coil. If we had gotten the 3 ton unit, we would have put a 15kW (rated) coil into the unit, however. We have 100A service to our house, which is fine for the 10kW heater especially given that our dryer and water heater both still run on natural gas. For the 15kW heater, we would probably still be fine with 100A service, since the heater operates at ~7kW in all but the very coldest conditions. That said, if we had needed to upgrade the electrical, it would be an extra $7000 or so upfront cost; we’ll presumably want to upgrade the electrical someday, so arguably this is an expense delayed, rather than an expense saved. Last time I checked, the voltage that we get is closer to 220V than 240V, which means the actual capacity of the heating coils is about 10% less than the rated capacity.
The challenge happens after -25C. It is very rare, but worth considering. The Mitsubishi may operate as low as -30C, but as far as I can tell from the fine print, it’s not really guaranteed to operate lower than -25C. There is a slight risk of a cold snap, certainly very rare, that drops the temperature below the operating temperature of the heat pump, leaving us with electrical resistance heating only. If the electrical resistance heater in the air handler cannot keep up, then we would have to supplement with space heaters plugged into the regular 15A (110V) electrical. Alternately, we could live with a colder house — not entirely unreasonable given that this is almost certainly less than a once-in-a-decade event. Furthermore, when the conditions do drop to colder than -25C, it almost certainly will be a night time, when it is okay for the house temperature to drop by several degrees anyways.
All the same, we do keep a number of space heaters in our home – my office space is unheated, so there are a couple space heaters in my office space. It’s a huge shock and death risk to keep a space heater in your shower + bathroom area, but I admit that we have one there, too (please don’t do this!). Trying to buy a space heater in the midst of a once-in-a-decade cold snap is a non-starter, so they need to be ready and waiting in our house in order to be of any use.
Comparing the Large and Small Mitsubishi Units
Let’s put that all in table, using a weighted hydro cost of $0.10 per kWh. Various bits of guesswork here, of course.
| Outside Temp | est. kW req’d | Small heat pump | Resistance heater w/ small HP | Large HP | Resistance heater w/ large HP | resistance heater difference | assumed COP for HP | $ difference per hour, at $0.1/kWh |
| -13C | 7kW | 7kW (rated 8kW) | 0kW | 7kW (rated 11kW) | 0kW | 0kW | 1.8 | $0 |
| -20C | 10kW | 6.5kW (guess) | 3.5kW | 9kW (guess) (*) | 1kW | 2.5kW | 1.65 (guess) | $0.10 |
| -25C | 13kW | 6kW (rated 7kW) | 7kW | 8kW (rated 9kW) | 5kW | 2kW | 1.5 | $0.07 |
| -30C | 15kW | 4.5kW (guess – if it still operates) | 9kW + 1.5kW from space heaters | 7kW (guess) | 8kW | 2.5kW | 1.3 (guess) | $0.06 |
| -30C | 15kW | 0 (if it stops operating) | 9kW + 6kW from space heaters!! | 0 (if it stops operating) | 14kW + 1kW from space heater | 0kW (but space heaters cost money to buy) | n/a | n/a |
(*) There is an asterisk here. Heat pumps do not heat the air to the same temperature as a gas furnace. It’s not that it’s technologically infeasible, but rather that it’s cheaper to push more air through lower-temperature heat exchange coils. One of the hvac guys we spoke had the opinion that our ductwork was too small to support the full 9-10kW of heat from the large Mitsubishi. WIth hindsight, I think he was probably wrong, but who knows.
So somewhere between -15C and -25C, it costs about $0.10 more per hour to use the small heat pump versus the large heat pump.
Now, how much more does the large HP cost? Accounting for the $5000 subsidy and on an after-tax basis (careful!), it would only be around $300-$400 more. We were quoted the 10kW heater for the large Mitsubishi, so add a bit more for the 15kW upgrade that we would have taken. As mentioned earlier, the effective cost also increases since the large unit would make us ineligible for a later subsidy on tentatively planned window replacements. So let’s call it around $1500 more. Hopefully the lifespan is 15 to 20 years, so let’s say 15 years since that makes the math very simple. The math is happy since the $0.10/hour lower operating cost for the large unit is easy to work with, and gives us 15000 hours before the break-even point between the small unit and the large unit. At 15000 hours over 15 years, that’s about 1000 hours per year. Jan and Feb are the main months that matter, so about 500 hours per month, or about 15-20 hours per day, let’s say. Although it’s hard to believe, the actual average temperature through the full course of a day in Jan/Feb is more like -5C, and pretty much always warmer than -10C. Of course, it’s -13C right now as I’m writing this … which highlights the fact that averages can be misleading, since one day of 30C weather followed by one day of -30C will have an average temperature of 0C but an actual hvac operating cost that is far greater than two days of 0C. All the same, I’m very confident that the temperature is -15C or colder for far fewer than 15 hours per average Jan/Feb day. So for us, the small Mitsubishi wins on a financial basis.
There are some quibbles. I suppose the larger HP might allow you to easily heat the house up when hydro is at a low cost of $0.07/kWh and then turn off the heat when hydro prices increasing (time-of-use pricing for Ontario, see https://www.oeb.ca/consumer-information-and-protection/electricity-rates). But this won’t change much. The real response here, of course, is: if you had truly wanted to minimize total cost, then you would have left the old gas furnace inplace (or at most, purchased a new gas furnace). Eventually, when forced to switch off of gas (a couple decades from now, perhaps?), it’s reasonable to assume that the cost of an electric heat pump will have come down dramatically (and they will probably be more efficient – there is lots of room to improve the defrost cycle, for example).
The Conclusion?
It doesn’t take too long to look at the numbers and realize that it’s a lot easier to heat a new modern house than an old leaky house. There are some easy things we could do – for example, since we have a gas water heater and gas dryer, we need to ventilate the hvac room, which is done with a 5cm diameter hole in the wall. I wouldn’t be surprised if there’s a couple hundred watts of heat lost right there. Perhaps we could also be more responsible parents? Simple things like teaching kids to properly close windows can help (yup, and now the outer pane of the double window is frozen open, turning it into effectively a single pane window with an (admittedly attractive) frost pattern on the inside). Actually, the frost patterns suggest that not all of our double-paned windows are fully sealed anyways. For that matter, how often does the dog need to go out and chase a squirrel anyways? All these add-up and increase the heating load. Of course, we didn’t really do any of this with the primary objective of saving money, so I guess the dog still gets to chase its squirrels.
The Unexpected Downsides
There were some drawbacks we hadn’t expected. First, the heat pump runs nearly fulltime (it simply varies the fan speed and compressor), whereas our old furnace would cycle on/off. More air circulation means more money spent on replacing furnace filters (although perhaps a bit less on vacuum cleaner bags). There is more noise than I had expected — even at low throttle, air flowing through the air ducts makes noise. At full throttle, the indoor unit can shake and vibrate, causing some additional noise (note that it will rarely run at full throttle if the electrical resistance heating is enabled). There’s also a bit of noise as the unit switches between heating and defrost modes (and surprisingly, this can be heard from the ductless head, even though the ductless head is fully turned off during the winter).
Very specific to the particular manufacturer we have: I was surprised to find that we cannot set the thermostat to less than 16C or 17C (depending on the programming mode). Not that we do this often, but if we’re away, we sometimes will.
A Caveat For Really Warm Homes
As mentioned in the (*) note, I believe that most presently-available heat pumps typically do not heat the air to the same temperature as a gas furnace. We’ll probably need to keep this in mind if we ever decide to bump-up our temperature to a cozy 23C or so. I can’t imagine we’ll want to do this, but… I strongly suspect that raising the temperature to 23C instead of 18C is a much bigger efficiency loss for a heat pump than for a furnace. If nothing else, you can envision the impact of having a set point that is progressively closer to the temperature of the air coming out of the ductwork: the air from the ducts is about 30-35C for our heat pump instead of 50-55C for our old furnace. Of course, I would imagine that eventually there will be more options for homes with limited ductwork that require output air to have a high temperature. Choice is still limited – in our case, only the Mitsubishi was small enough to even fit into the available outdoor space.
Is there a recommendation?
For us, given the desire to improve the upstairs summer cooling situation and the tentative plan to use part of the $5000 subsidy for window replacements, the smaller Mitsubishi seems to have made sense (financially, anyways). For most people, if you are not planning to use a subsidy, then Mitsubishi has another unit that probably makes more sense than either of the two we considered. If there is no need for a ductless cooling head, and the full subsidy is in-play (and only in-play for the large Mitsubishi), then probably the large unit makes the most sense.
The Details
The large Mitsubishi we considered was the puz-ha36nka outdoor unit (https://www.mitsubishitechinfo.ca/sites/default/files/SB_PVA-A36AA7_PUZ-HA36NKA_202103.pdf) plus the pva-a36aa7 air handler. The small Mitsubishi was the mxz-3c30nahz3 outdoor unit (https://www.mitsubishitechinfo.ca/sites/default/files/SB_MXZ-3C30NAHZ3-U1_202207.pdf), combined with the svz-kp30na air handler and msz-gl06na ductless head. If we didn’t want the ductless head and we were (for some reason) ineligible for the $5000 subsidy, then the suz-ka36nahz (https://www.mitsubishitechinfo.ca/sites/default/files/SB_SVZ-KP36NA%20_SUZ-KA36NAHZ_202210.pdf) would have been interesting for us to look at.
For someone who really wants to be smart, there is room to take advantage of TOU hydro pricing. The peak price and off-peak price vary by about 2X in Ontario (https://www.oeb.ca/consumer-information-and-protection/electricity-rates). The COP of the heat pump varies from about 1.8 downwards (depending on the outside temperature). I’ve assumed heat loss from our house increases by a bit less than 0.5kW per degree difference between interior vs exterior temperature. There’s room here to optimize, for someone who really wants to. A smarter thermostat would figure it all out on its own.
